Question

import multiprocessing as mlp

class Test(mlp.Process):
    def run(self):
        if self.name == "Test-2":
            import time
            time.sleep(4)
        print self.name

jobs = []
for i in range(5):
    m = Test()
    jobs.append(m)
    m.run()
for i in jobs:
    i.join()

In the above code, output is always:
Test-1
Test-2
Test-3
Test-4
Test-5

And these processes spawn only one after another. Like Test-3 spans only after Test-2 is completed. Shouldn’t they happen parallely???

And also, i.join() throws error: AssertionError: can only join a started process

What wrong am I doing???

Thanks…

Answer 1

You never started the process. The run method is intended for the subclass to implement, which you correctly did, but the start method is the one you should call to start the process (and automatically execute run() in the other process instance).

In other words, call m.start() instead of m.run().